This research will investigate the acquisition of American Sign Language (ASL), a natural language of the deaf of North America. ASL may be acquired as a primary language either early or late in life, and either from native signers or from signers who themselves acquired the language late in life and have had little exposure to it. Its study therefore provides an unusual opportunity to investigate the consequences of early vs. late experience, and of wide variations in input environment, on the process by which a language is acquired and the ultimate character of its users' knowledge. These questions are theoretically significant and also crucial for applied decisions surrounding the lives of deaf children. Previous work has focused primarily on early vs. late exposure, showing that age of exposure strongly influences performance, many years later, in both a first and second language, and on both language-specific and language universal aspects of language. Previous work also has included some studies of the consequences of moderately impoverished input environments on early language acquisition. This work has shown that early learners of a language are able to reorganize inconsistent aspects of their input and thereby surpass their own input models, acquiring a language rather different and more well-structured than the one to which they were exposed. In the present application, five studies are proposed, all examining the consequences of more severely impoverished input environments. Two studies investigate the longitudinal acquisition of ASL by children exposed to either moderately or severely impoverished input, analyzing both the input sources and the children's linguistic competence to see whether, and by what means, children may also surpass these more degraded input sources. Two studies examine the ultimate outcome in adulthood, of such early exposure. One final study explores the possibility of simulating such acquisition processes in a laboratory paradigm. The results should contribute to our understanding of the linguistic consequences of varying types of early input for language acquisition, and to the character and processes of language learning in childhood. In addition, they should contribute to decisions regarding quality of early language exposure, whether spoken or signed, in deaf education and parent counseling.